An underwater vehicle attitude inversion method fusing artificial lateral line and gyroscope

Abstract

The present application relates to the field of pressure signal model, specifically relates to a kind of underwater vehicle attitude inversion method of fusion artificial lateral line and gyroscope, including 1 constructs underwater vehicle physical model and space coordinate system, determines motion parameter and flow field environmental parameter;2 based on the lateral line perception principle of fish, constructs artificial lateral line array;3 according to attitude motion characteristics, constructs underwater vehicle attitude representation method;4 according to the characteristics of fluid mechanics, constructs underwater vehicle kinematic pressure signal model;5 design experimental prototype, carry out pool experiment to generate data set, pre-process signal;6 constructs residual objective function, carries out attitude inversion to underwater vehicle;7 the relative root mean square error is used to evaluate inversion accuracy, completes the reliability verification of the above attitude inversion method.The present application constructs the kinematic pressure signal model of underwater vehicle based on quaternion model by fusion artificial lateral line and gyroscope, improves underwater vehicle attitude inversion precision and stability.

Description

Technical Field

[0001] This invention relates to the field of multi-parameter kinematic pressure signal models for underwater vehicles, and specifically to a method for attitude inversion of underwater vehicles that integrates artificial lateral lines and gyroscopes. Background Technology

[0002] Lateral line is one of the main organs by which fish perceive information about the surrounding flow field, effectively ensuring that fish can successfully complete complex movements such as foraging and obstacle avoidance. Traditional inertial sensing systems rely on integral calculations to calculate the attitude of underwater vehicles, a process that inevitably introduces accumulated errors. However, gyroscopes have the advantage of high short-term accuracy and are the only direct source for calculating the attitude of underwater vehicles. In contrast, the pressure signal measured by artificial lateral line contains instantaneous motion information of the underwater vehicle, effectively avoiding the interference of accumulated errors generated by the integral calculation process, and providing a new approach to realize the attitude inversion of underwater vehicles. Therefore, establishing a kinematic pressure signal model of underwater vehicles based on artificial lateral line, integrating gyroscope angular velocity measurements, and realizing the attitude inversion of underwater vehicles to improve the navigation and positioning accuracy of underwater vehicles is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0003] To address the problems existing in the prior art, the purpose of this invention is to overcome the shortcomings of the cumulative error generated during the integration process of traditional inertial sensing systems, so as to improve the motion control accuracy of underwater vehicles. Based on the principle of fish lateral line perception, this invention proposes an underwater vehicle attitude inversion method that integrates artificial lateral line and gyroscope. A kinematic pressure signal model of the underwater vehicle is established based on the artificial lateral line, and the instantaneous angular velocity measurement value of the gyroscope is fused to realize the attitude inversion of the underwater vehicle.

[0004] To achieve the above objectives, the technical solution adopted by this invention is: a method for attitude inversion of underwater vehicles that integrates artificial lateral line and gyroscope, comprising the following steps:

[0005] Step 1: Construct the physical model and spatial coordinate system of the underwater vehicle, describe the relationship between the motion parameters of the underwater vehicle and the flow field environment parameters, and determine the motion parameters and flow field environment parameters of the vehicle;

[0006] Step 2: Based on the principle of lateral line perception in fish, pressure sensors are selected to construct an artificial lateral line array, and the array layout and the location of the pressure sensors are determined.

[0007] Step 3: Based on the attitude and motion characteristics of the underwater vehicle, construct an underwater vehicle attitude representation method based on a quaternion model;

[0008] Step 4: Based on the hydrostatic and dynamic characteristics, construct a multi-parameter kinematic pressure signal model for the underwater vehicle;

[0009] Step 5: Design an experimental prototype based on the physical model parameters of the underwater vehicle. The artificial lateral pressure array is mounted externally to collect flow field pressure signals. The gyroscope unit is installed at the center of mass of the experimental prototype to collect instantaneous angular velocity in the body coordinate system. A water tank experiment is conducted to generate a dataset. The collected pressure signals and gyroscope signals are preprocessed.

[0010] Step 6: Fuse the preprocessed instantaneous angular velocity measurements from the gyroscope as supplementary observation information, and construct a residual objective function using the underwater vehicle's motion parameters and flow field environment parameters as parameters to be estimated, to perform attitude inversion on the underwater vehicle;

[0011] Step 7: Compare the obtained underwater vehicle attitude inversion results with the actual attitude information, use relative root mean square error to evaluate the inversion accuracy, and complete the reliability verification of the above attitude inversion method.

[0012] In the above-mentioned underwater vehicle attitude inversion method that integrates artificial lateral line and gyroscope, in step 1, the underwater vehicle is a torpedo-type underwater vehicle, the spatial coordinate system includes three types of coordinate systems: the Earth fixed coordinate system, the navigation coordinate system, and the body coordinate system. The motion parameters of the underwater vehicle include attitude parameters in the navigation coordinate system, speed and angular velocity parameters in the body coordinate system, and flow field environment parameters are defined as flow field velocity parameters in the navigation coordinate system.

[0013] In the above-mentioned underwater vehicle attitude inversion method that integrates artificial sideline and gyroscope, in step 2, the torpedo-shaped underwater vehicle adopts an axisymmetric structure, the pressure sensor extends along the axial direction of the torpedo-shaped underwater vehicle and is arranged circumferentially, and the core layout parameters of the artificial sideline array include the axial spacing distance of the main array elements, the circumferential spacing angle of the head array elements, and the centroid offset distance.

[0014] The above-mentioned underwater vehicle attitude inversion method that integrates artificial lateral line and gyroscope, in step 3, the quaternion model q By a rotation angle α and a unit rotation axis u The components of the unit rotation axis in the navigation coordinate system are ( u x , u y , u z In the navigation coordinate system, any attitude transformation of an underwater vehicle is equivalent to the body coordinate system rotating around a unit axis. u Rotation angle α Based on the above geometric relationships, step 3 includes:

[0015] Step 3-1: Define the basic form of the quaternion model describing the vehicle's attitude:

[0016] ,in: q 0 is the real part of the quaternion. q 1. q 2. q 3 is the imaginary part of the quaternion, sin is the sine function, and cos is the cosine function;

[0017] Step 3-2: Derive the rotation matrix using the quaternion model C b n The matrix form of the quaternion model can be represented as:

[0018] ;

[0019] Step 3-3: Determine the coordinates M of any pressure sensor of the underwater vehicle in the body coordinate system. i,j b The coordinates M of the underwater vehicle in the navigation coordinate system after it rotates. i,j n The expression: .

[0020] The above-mentioned underwater vehicle attitude inversion method that integrates artificial lateral line and gyroscope, step 4 includes:

[0021] Step 4-1: Based on the fact that the measured values ​​of any pressure sensor in the artificial lateral line array are the result of the superposition of hydrostatic pressure and multi-source dynamic pressure, the overall expression of the kinematic pressure signal model of the underwater vehicle is obtained as follows: ,in: p i,j For the first i The first axis, the first j Pressure measurement value from a circumferential pressure sensor p s,i,j For the first i The first axis, the first j The static pressure of a circumferential pressure sensor p d,i,j This is the sum of the dynamic pressure generated by the flow field, speed, and angular velocity at the pressure sensor.

[0022] Step 4-2: Based on the fundamental properties of hydrostatics, and combining the Earth's fixed coordinate system and the navigation coordinate system, determine the water depth at which the underwater vehicle's mass point is located. d 0, and the coordinates of each sensor on the artificial lateral line in the navigation coordinate system, the hydrostatic model can be expressed as: ,in: p 0 represents atmospheric pressure. r The density of seawater, g For gravitational acceleration, (M i,j n)3 is the pressure sensor in the navigation coordinate system z Coordinate values ​​along the axis;

[0023] Step 4-3: Determine the multi-source dynamic pressure based on the velocity dynamic pressure component, the speed dynamic pressure component, and the angular velocity dynamic pressure component. Superimpose the static pressure with each dynamic pressure component of the multi-source dynamic pressure to obtain the multi-parameter kinematic pressure signal model of the artificial flank array.

[0024] The above-mentioned underwater vehicle attitude inversion method that integrates artificial lateral line and gyroscope, wherein step 4-3 includes:

[0025] A: According to fluid dynamics characteristics, the flow velocity and dynamic pressure are determined by the normal velocity of the ambient flow field relative to the pressure sensor. The flow velocity parameters of the horizontal flow field in the navigation coordinate system are... v f n for: ,in: v fx n and v fy n The flow field in the navigation coordinate system is respectively x Axial direction and y The component along the axis, T is the transpose operation;

[0026] The horizontal flow field in the body coordinate system is equivalent to the flow field about the unit rotation axis in the navigation coordinate system. u Rotation Angle - α The rotation matrix at this point is C n b Then the velocity parameters of the horizontal flow field in the body coordinate system v f b Represented as: ;

[0027] The normal velocity of the horizontal flow field relative to the pressure sensor is the projection of the flow field velocity onto the unit normal vector of the sensor. v f,i,j b Confirmed, expressed as: The velocity-dynamic pressure model is then expressed as: ,in: p d,f,i,j For the first i The first axis, the first j The flow velocity dynamic pressure of a circumferential pressure sensor;

[0028] b: Based on the vehicle's motion characteristics, the speed dynamic pressure is determined by the vehicle's forward motion relative to the pressure sensor's normal velocity, and the speed parameters in the body coordinate system. vv b Represented as: ,in: v vx b For the airspeed in the body coordinate system x Components along the axial direction;

[0029] The forward velocity of the aircraft relative to the pressure sensor is the projection of the aircraft's speed onto the unit normal vector of the sensor. v v,i,j b Confirmed, expressed as: The speed dynamic pressure model is then expressed as: ,in: p d,v,i,j For the first i The first axis, the first j The speed dynamic pressure of a circumferential pressure sensor;

[0030] c: Based on the aircraft's rotational characteristics, the angular velocity dynamic pressure is determined by the aircraft's own rotational motion relative to the normal velocity of the pressure sensor, and the angular velocity parameters in the body coordinate system. w b Represented as: ,in: w x b , w y b , w z b Angular velocity in body coordinate system x Axial direction, y Axial direction and z Components along the axial direction;

[0031] Combining the coordinates of each sensor along the artificial lateral line in the body coordinate system, the linear velocity at each sensor location caused by the spacecraft's own rotational motion. v w b Represented as: , of which: | w b | represents the magnitude of the angular velocity parameter;

[0032] The normal velocity of the aircraft's rotational motion relative to the pressure sensor is the projection of the linear velocity onto the unit normal vector of the sensor. v w,i,j b Confirmed, expressed as: Then the angular velocity dynamic pressure model is expressed as: ,in: p d,w,i,j For the firsti The first axis, the first j The angular velocity dynamic pressure of a circumferential pressure sensor.

[0033] The above-mentioned underwater vehicle attitude inversion method that integrates artificial sideline and gyroscope, in step 5, selects a hydrodynamic experimental pool to simulate the horizontal environmental flow field, designs multi-dimensional experimental conditions, realizes complex motion control of the experimental prototype through a six-degree-of-freedom experimental platform, places the experimental prototype in the flow field, collects static data for sensor calibration, and simultaneously collects artificial sideline array pressure signals and gyroscope angular velocity signals to generate datasets.

[0034] The initial offset of the sensor is eliminated by using the mean of static pressure data as the baseline, and the inherent zero bias of the gyroscope is eliminated by using the mean of static angular velocity data as the zero bias value. The pressure signal and angular velocity signal are filtered separately to remove high-frequency turbulence noise and suppress random noise interference.

[0035] The above-mentioned underwater vehicle attitude inversion method that integrates artificial lateral line and gyroscope, step 6 includes:

[0036] Step 6-1: Define the parameter vector to be estimated, with motion parameters and flow field parameters as the core. i for: The preprocessed instantaneous angular velocity of the gyroscope is used as known observation information and directly fused into the multi-parameter kinematic pressure signal model constructed in step 4 as the input of the angular velocity dynamic pressure component.

[0037] Step 6-2: Using the preprocessed artificial lateral line array pressure signal as the measured value and the multi-parameter kinematic pressure signal model that integrates gyroscope observation information as the theoretical value, construct the residual objective function. J ( i ), represented as: Where: A and B are the number of axial and circumferential sensors in the artificial side-line array in step 2, respectively. p i,j ( t )for t Measured values ​​of artificial side-line array pressure signals after time-lapse preprocessing. p i,j ( i The theoretical pressure value is obtained by substituting the parameters to be estimated into the multi-parameter kinematic pressure signal model.

[0038] Step 6-3: Iterative process of attitude inversion based on LM algorithm, determining the initial values ​​of the parameters to be estimated. i 0. Convergence threshold e and initial value of damping coefficient m 0. Calculate the theoretical pressure values ​​corresponding to all pressure sensors, and calculate the Jacobian matrix of the residual objective function with respect to each parameter to be estimated.J k Calculate the parameter correction amount according to the LM algorithm formula. Dth k : ,in: k This represents the number of iterations in the LM algorithm. m k The damping coefficient is... I It is the identity matrix. r k The residual vector;

[0039] Step 6-4: Update the parameters to be estimated and calculate the updated residual objective function value. Determine whether the iteration has converged based on the convergence threshold. If it has not converged, adjust the damping coefficient based on the residual objective function value and continue iterating until the iteration converges. Extract quaternions from the optimal parameters to achieve attitude inversion.

[0040] The aforementioned underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope, in step 7: using the actual attitude information of the six-degree-of-freedom experimental platform used in the pool experiment as a benchmark, and converting it into quaternion form, the attitude inversion results output in step 6 are time-series matched with the actual attitude data. The accuracy of the underwater vehicle attitude inversion is quantitatively characterized by the relative root mean square error of the four attitude parameters of the quaternion model. The relative root mean square error per unit rotation axis of the quaternion model is expressed as... The relative root mean square error of the rotation angle of the quaternion model can be expressed as: , where: RRMSE( u ) represents the relative root mean square error (RRMSE) of attitude inversion per unit rotation axis. α ) represents the relative root mean square error of the rotation angle attitude inversion. t Sampling time, l The number of sampling points under experimental conditions. u 1( t )for t The rotation axis obtained by attitude inversion at any given time. u 0( t )for t Real-time rotation axis α 1( t )for t The rotation angle obtained from the attitude inversion at any given moment. α 0( t )for t Real-time rotation angle.

[0041] The beneficial effects of this invention's underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope are as follows: This invention constructs an underwater vehicle kinematic pressure signal model based on a quaternion model by integrating artificial lateral line and gyroscope. It leverages the advantage of the quaternion model's absence of singularities in its computation process to compensate for the gimbal lock deficiency of the traditional Euler angle model, fully utilizing the high short-time accuracy of the gyroscope while avoiding the cumulative error introduced by its integration process. Combining the physical model structural characteristics of a torpedo-type underwater vehicle, an artificial lateral line pressure sensor array is constructed based on the principle of fish lateral line flow field perception. A multi-parameter kinematic pressure signal model is established by combining hydrostatic and dynamic characteristics. Finally, a pressure dataset is generated through pool experiments, and gyroscope signals are integrated to supplement observation information. A residual objective function is constructed for attitude inversion, and root mean square error is used to evaluate the attitude inversion accuracy. This forms a complete method verification technical solution, significantly reducing the design cost and energy consumption of the navigation system, improving the accuracy and stability of underwater vehicle attitude inversion, and laying a technical foundation for the scientific navigation and positioning of underwater vehicles. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the physical model and spatial coordinate system of the torpedo-type underwater vehicle in an embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of the artificial lateral pressure sensor array of an underwater vehicle in an embodiment of the present invention;

[0044] Figure 3 This is a schematic diagram of the quaternion model attitude representation method in an embodiment of the present invention;

[0045] Figure 4 This is a flowchart of the attitude inversion process of the LM algorithm in an embodiment of the present invention;

[0046] Figure 5 This is a flowchart of the attitude inversion method that integrates artificial lateral line and gyroscope in an embodiment of the present invention. Detailed Implementation

[0047] To enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention will be described below in conjunction with specific embodiments and accompanying drawings.

[0048] like Figure 5 As shown, an underwater vehicle attitude inversion method that integrates artificial lateral line and gyroscope includes the following steps.

[0049] S1. Construct the physical model and spatial coordinate system of the underwater vehicle, and determine the motion and flow field environment parameters of the vehicle.

[0050] like Figure 1As shown, this embodiment takes a torpedo-type underwater vehicle as the research object, and simplifies its structural features to a hemispherical head and a cylindrical body.

[0051] This embodiment defines three types of coordinate systems: the Earth fixed coordinate system, the navigation coordinate system, and the body coordinate system, to describe the relationship between the motion parameters of the underwater vehicle and the flow field environment. All three types of coordinate systems are right-handed rectangular coordinate systems.

[0052] Earth fixed coordinate system O e - x e y e z e As a global reference frame, used to describe the relative position changes of the underwater vehicle, the origin of the coordinate system is selected as the projection point of the underwater vehicle's initial position centroid onto the sea level. O e , x e The axis extends northward. y e The axis extends eastward. z e The axis points vertically downwards along the Earth's center.

[0053] Navigation coordinate system O n - x n y n z n Used to visually represent the vehicle's position, attitude information, and flow field parameters; coordinate system origin. O n It is fixed to the center of mass of the spacecraft, and its coordinate axes are aligned with the Earth's fixed coordinate system.

[0054] Body coordinate system O b - x b y b z b Used to describe the relative motion parameters of the aircraft itself; origin of the coordinate system. O b Also fixed to the center of mass of the spacecraft x b The axis is along the direction of the vehicle's movement. y b The axis points laterally to starboard along the side of the aircraft. z b Axis perpendicular to x bO b y b The plane points towards the ventral side of the aircraft.

[0055] Based on the above coordinate system, the motion parameters of the aircraft and the flow field environment parameters are defined. The motion parameters of the aircraft include attitude parameters in the navigation coordinate system and speed and angular velocity parameters in the body coordinate system. The flow field environment parameters are defined as the flow field velocity parameters in the navigation coordinate system.

[0056] S2. Based on the principle of fish lateral line perception, pressure sensors are selected to construct an artificial lateral line array, and the array layout and the position of the pressure sensors are determined.

[0057] Fish rely on their lateral line, with its canal nerve tumulus, to keenly sense changes in water pressure. Combining this with the axisymmetric structure of torpedo-shaped underwater vehicles, pressure sensors are extended axially and arranged circumferentially. Figure 2 As shown in the diagram, this layout can synchronously sense changes in the pressure field caused by different motion states of the aircraft, providing a reliable sensing platform for subsequent pressure signal model construction.

[0058] The core parameters of artificial side-line array layout include the axial spacing of the main array elements. Δd Circumferential spacing angle of the head array element β and centroid offset distance d .

[0059] Using the body coordinate system O b - x b y b z b Based on this, determine the first i The first axis, the first j The three-dimensional coordinates M of a circumferential pressure sensor i,j b The normal vector of the pressure sensor is the direction of the outward normal to the surface of the aircraft, determining the first... i The first axis, the first j The unit normal vector N of the circumferential pressure sensor i,j b .

[0060] S3, Construct an underwater vehicle attitude representation method based on a quaternion model.

[0061] This embodiment focuses on the attitude and motion characteristics of a torpedo-type underwater vehicle, using a quaternion model. q It is a four-dimensional model that does not exhibit singularity issues during large-angle attitude changes. Quaternion model q By a rotation angle αand a unit rotation axis u The components of the unit rotation axis in the navigation coordinate system are ( u x , u y , u z ),like Figure 3 As shown, arbitrary attitude transformation of an underwater vehicle in the navigation coordinate system is equivalent to the body coordinate system rotating around a unit axis. u Rotation angle α .

[0062] Based on the above geometric relationships, the basic form of the quaternion model describing the attitude of an aircraft is defined as follows:

[0063] (1), where: q 0 is the real part of the quaternion. q 1. q 2. q 3 is the imaginary part of the quaternion and satisfies the unit constraint of the quaternion. q 0 2 + q 1 2 + q 2 2 + q 3 2 =1, sin is the sine function, and cos is the cosine function.

[0064] To achieve the transformation of physical quantities between the navigation coordinate system and the body coordinate system, the rotation matrix needs to be derived using a quaternion model. C b n The matrix form of the quaternion model can be represented as:

[0065] (2).

[0066] For any pressure sensor coordinate M of the underwater vehicle in the body coordinate system i,j b The coordinates M of the underwater vehicle in the navigation coordinate system after it rotates. i,j n It can be expressed by the following formula:

[0067] (3).

[0068] Formula (3) can be obtained by multiplying a matrix and a vector, and is used to obtain the coordinates of each sensor in the navigation coordinate system.

[0069] S4. Based on hydrostatic and dynamic characteristics, a multi-parameter kinematic pressure signal model of an underwater vehicle is constructed.

[0070] The measurement value of any pressure sensor in the artificial lateral line array is the result of the superposition of hydrostatic pressure and multi-source dynamic pressure. The overall expression of the kinematic pressure signal model of the underwater vehicle is:

[0071] (4), where: p i,j For the first i The first axis, the first j Pressure measurement value from a circumferential pressure sensor p s,i,j This is the hydrostatic pressure at the pressure sensor. p d,i,j It is the sum of the dynamic pressure generated by the flow field, speed, and angular velocity at the pressure sensor.

[0072] Based on the fundamental properties of hydrostatics, hydrostatic pressure is determined by the water depth where the sensor is located and is isotropic. The water depth of the underwater vehicle's particle is determined by combining the Earth's fixed coordinate system and the navigation coordinate system. d 0. Combining the coordinates of each sensor on the artificial lateral line in the navigation coordinate system, the hydrostatic model can be expressed as:

[0073] (5), where: p 0 represents atmospheric pressure. r The density of seawater, g For gravitational acceleration, (M i,j n )3 is the pressure sensor in the navigation coordinate system z The coordinate values ​​along the axis.

[0074] Multi-source dynamic pressure consists of velocity dynamic pressure components, speed dynamic pressure components, and angular velocity dynamic pressure components. According to fluid dynamics characteristics, velocity dynamic pressure is determined by the normal velocity of the ambient flow field relative to the pressure sensor. The velocity parameters of the horizontal flow field in the navigation coordinate system are also considered. v f n for:

[0075] (6), where: v fx n and v fy n The flow field in the navigation coordinate system is respectively x Axial direction and y The component along the axis, T is the transpose operation.

[0076] The horizontal flow field in the body coordinate system is equivalent to the flow field about the unit rotation axis in the navigation coordinate system. u Rotation Angle -α The rotation matrix at this point is C n b Then the velocity parameters of the horizontal flow field in the body coordinate system v f b It can be represented as:

[0077] (7).

[0078] The normal velocity of the horizontal flow field relative to the pressure sensor is the projection of the flow field velocity onto the unit normal vector of the sensor. v f,i,j b Determined can be represented as:

[0079] (8).

[0080] The velocity-pressure model can then be expressed as:

[0081] (9), where: p d,f,i,j For the first i The first axis, the first j The flow velocity dynamic pressure of a circumferential pressure sensor.

[0082] Based on the vehicle's motion characteristics, the dynamic pressure of the airspeed is determined by the normal velocity of the vehicle's forward motion relative to the pressure sensor. Airspeed parameters in the body coordinate system. v v b It can be represented as:

[0083] (10), where: v vx b For the airspeed in the body coordinate system x Components in the axial direction.

[0084] The forward velocity of the aircraft relative to the pressure sensor is the projection of the aircraft's speed onto the unit normal vector of the sensor. v v,i,j b Determined can be represented as:

[0085] (11).

[0086] The speed dynamic pressure model can then be expressed as:

[0087] (12), where: p d,v,i,j For the first i The first axis, the firstj The speed dynamic pressure of a circumferential pressure sensor.

[0088] Based on the aircraft's rotational characteristics, the angular velocity dynamic pressure is determined by the aircraft's own rotational motion relative to the normal velocity of the pressure sensor. Angular velocity parameters in the body coordinate system. w b It can be represented as:

[0089] (13), where: w x b , w y b , w z b Angular velocity in body coordinate system x Axial direction, y Axial direction and z Components in the axial direction.

[0090] Combining the coordinates of each sensor along the artificial lateral line in the body coordinate system, the linear velocity at each sensor location caused by the spacecraft's own rotational motion. v w b It can be represented as:

[0091] (14), where: | w b | represents the magnitude of the angular velocity parameter.

[0092] The normal velocity of the aircraft's rotational motion relative to the pressure sensor is the projection of the linear velocity onto the unit normal vector of the sensor. v w,i,j b Determined can be represented as:

[0093] (15).

[0094] The angular velocity dynamic pressure model can then be expressed as:

[0095] (16), where: p d,w,i,j For the first i The first axis, the first j The angular velocity dynamic pressure of a circumferential pressure sensor.

[0096] By superimposing the above static pressure with each dynamic pressure component, a multi-parameter kinematic pressure signal model of the artificial lateral line array is obtained.

[0097] S5 performs a water tank experiment to generate a dataset and preprocesses the collected pressure signals and gyroscope signals.

[0098] This embodiment designs an experimental prototype based on the physical model parameters of a torpedo-type underwater vehicle. Externally, it is equipped with the artificial lateral pressure sensor array designed in S2 to collect pressure signals from the guiding flow field. Internally, a gyroscope unit is installed at the vehicle's center of mass. O b This is used to collect instantaneous angular velocity in the body coordinate system.

[0099] A hydrodynamic experimental tank was selected to simulate a horizontal flow field. Multi-dimensional experimental conditions were designed, and complex motion control of the experimental prototype was achieved through a six-degree-of-freedom experimental platform. The experimental prototype was placed statically in the flow field, and static data was collected for sensor calibration. Simultaneously, pressure signals from an artificial lateral line array and angular velocity signals from a gyroscope were collected to generate a dataset.

[0100] To address issues such as noise and zero bias in the original signal, the initial offset of the sensor is eliminated by using the mean of static pressure data as the baseline, and the inherent zero bias of the gyroscope is eliminated by using the mean of static angular velocity data as the zero bias value. The pressure signal and angular velocity signal are filtered separately to remove high-frequency turbulence noise and suppress random noise interference.

[0101] S6 integrates the instantaneous angular velocity measurements from the gyroscope after preprocessing in S5 as supplementary observation information. Using the motion parameters of the underwater vehicle and the flow field environment parameters as parameters to be estimated, a residual objective function is constructed to perform attitude inversion on the underwater vehicle.

[0102] The pose inversion process of the LM algorithm is as follows: Figure 4 As shown.

[0103] With motion parameters and flow field parameters as the core, a parameter vector to be estimated is defined. i for:

[0104] (17).

[0105] The instantaneous angular velocity of the gyroscope preprocessed by S5 is used as known observation information and directly fused into the multi-parameter kinematic pressure signal model constructed by S4 as the input of the angular velocity dynamic pressure component, thereby reducing the dimensionality of the parameters to be estimated and the computational complexity.

[0106] Using the preprocessed artificial lateral line array pressure signal from S5 as the measured value and the multi-parameter kinematic pressure signal model incorporating gyroscope observations as the theoretical value, a residual objective function is constructed. J ( i Its basic form can be expressed as:

[0107] (18), where: A and B are the number of axial and circumferential sensors in the artificial side-line array of S2, respectively. p i,j ( t )for t Measured values ​​of artificial side-line array pressure signals after time-lapse preprocessing. p i,j ( i The theoretical pressure value is obtained by substituting the parameters to be estimated into the multi-parameter kinematic pressure signal model.

[0108] The attitude inversion iterative process based on the LM algorithm first involves determining the initial values ​​of the parameters to be estimated. i 0. Convergence threshold e and initial value of damping coefficient m 0, calculate the corresponding theoretical pressure values ​​for all pressure sensors.

[0109] Calculate the Jacobian matrix of the residual objective function with respect to each parameter to be estimated. J k Calculate the parameter correction amount according to the LM algorithm formula. Dth k :

[0110] (19), of which: k This represents the number of iterations in the LM algorithm. m k The damping coefficient is... I It is the identity matrix. r k This is the residual vector.

[0111] Finally, the parameters to be estimated are updated and the updated residual objective function value is calculated. The convergence threshold is used to determine whether the iteration has converged. If it has not converged, the damping coefficient is adjusted according to the residual objective function value, and the iteration continues until the iteration converges. Quaternions are extracted from the optimal parameters to achieve attitude inversion.

[0112] S7. Compare the obtained underwater vehicle attitude inversion results with the actual attitude information, and use the relative root mean square error to evaluate the inversion accuracy, thus completing the reliability verification of the above attitude inversion method.

[0113] This embodiment uses the actual attitude information of the six-degree-of-freedom experimental platform used in the pool experiment as a benchmark, and converts it into quaternion form. The attitude inversion results output by S6 are matched with the actual attitude data over time. The attitude inversion accuracy of the underwater vehicle can be quantitatively characterized by the relative root mean square error of the four attitude parameters of the quaternion model. The relative root mean square error of the unit rotation axis and rotation angle of the quaternion model can be expressed as follows:

[0114] (20)

[0115] (21), where: RRMSE( u ) represents the relative root mean square error (RRMSE) of attitude inversion per unit rotation axis. α ) represents the relative root mean square error of the rotation angle attitude inversion. t Sampling time, l The number of sampling points under experimental conditions. u 1( t )for t The rotation axis obtained by attitude inversion at any given time. u 0( t )for t Real-time rotation axis α 1( t )for t The rotation angle obtained from the attitude inversion at any given moment. α 0( t )for t Real-time rotation angle.

[0116] The relative root mean square error of the unit rotation axis and rotation angle of the quaternion model is used as the evaluation index for the attitude inversion accuracy of the artificial side-line array with fused gyroscope. The mean and maximum values ​​of the relative root mean square error of the underwater vehicle attitude inversion under all working conditions are statistically analyzed. The reliability of the above attitude inversion method is verified by using the threshold of actual engineering application as the judgment standard.

[0117] The above embodiments are merely illustrative of the structural concept and features of the present invention, intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly, and should not be construed as limiting the scope of protection of the present invention. All equivalent changes or modifications made based on the essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for attitude inversion of an underwater vehicle integrating artificial lateral line and gyroscope, characterized in that, Includes the following steps: Step 1: Construct the physical model and spatial coordinate system of the underwater vehicle, describe the relationship between the motion parameters of the underwater vehicle and the flow field environment parameters, and determine the motion parameters and flow field environment parameters of the vehicle; Step 2: Based on the principle of lateral line perception in fish, pressure sensors are selected to construct an artificial lateral line array, and the array layout and the location of the pressure sensors are determined. Step 3: Based on the attitude and motion characteristics of the underwater vehicle, construct an underwater vehicle attitude representation method based on a quaternion model; Step 4: Based on the hydrostatic and dynamic characteristics, construct a multi-parameter kinematic pressure signal model for the underwater vehicle. Determine the multi-source dynamic pressure based on the velocity dynamic pressure component, the speed dynamic pressure component, and the angular velocity dynamic pressure component. Superimpose the static pressure with the dynamic pressure components of the multi-source dynamic pressure to obtain the multi-parameter kinematic pressure signal model of the artificial flank array, including: A: According to fluid dynamics characteristics, the flow velocity and dynamic pressure are determined by the normal velocity of the ambient flow field relative to the pressure sensor. The flow velocity parameters of the horizontal flow field in the navigation coordinate system are... v f n for: ,in: v fx n and v fy n The flow field in the navigation coordinate system is respectively x Axial direction and y The component along the axis, T is the transpose operation; The horizontal flow field in the body coordinate system is equivalent to the flow field about the unit rotation axis in the navigation coordinate system. u Rotation angle - α The rotation matrix at this point is C n b Then the velocity parameters of the horizontal flow field in the body coordinate system v f b Represented as: ; The normal velocity of the horizontal flow field relative to the pressure sensor is the projection of the flow field velocity onto the unit normal vector of the sensor. v f,i,j b Confirmed, expressed as: The velocity-dynamic pressure model is then expressed as: ,in: p d,f,i,j For the first i The first axis, the first j The flow velocity dynamic pressure of a circumferential pressure sensor; b: Based on the vehicle's motion characteristics, the speed dynamic pressure is determined by the vehicle's forward motion relative to the pressure sensor's normal velocity, and the speed parameters in the body coordinate system. v v b Represented as: ,in: v vx b For the airspeed in the body coordinate system x Components along the axial direction; The forward velocity of the aircraft relative to the pressure sensor is the projection of the aircraft's speed onto the unit normal vector of the sensor. v v,i,j b Confirmed, expressed as: The speed dynamic pressure model is then expressed as: ,in: p d,v,i,j For the first i The first axis, the first j The speed dynamic pressure of a circumferential pressure sensor; c: Based on the aircraft's rotational characteristics, the angular velocity dynamic pressure is determined by the aircraft's own rotational motion relative to the normal velocity of the pressure sensor, and the angular velocity parameters in the body coordinate system. w b Represented as: ,in: w x b , w y b , w z b Angular velocity in body coordinate system x Axial direction, y Axial direction and z Components along the axial direction; Combining the coordinates of each sensor along the artificial lateral line in the body coordinate system, the linear velocity at each sensor location caused by the spacecraft's own rotational motion. v w b Represented as: , of which: | w b | represents the magnitude of the angular velocity parameter; The normal velocity of the aircraft's rotational motion relative to the pressure sensor is the projection of the linear velocity onto the unit normal vector of the sensor. v w,i,j b Confirmed, expressed as: Then the angular velocity dynamic pressure model is expressed as: ,in: p d,w,i,j For the first i The first axis, the first j Angular velocity dynamic pressure of a circumferential pressure sensor; Step 5: Design an experimental prototype based on the physical model parameters of the underwater vehicle. The artificial lateral pressure array is mounted externally to collect flow field pressure signals. The gyroscope unit is installed at the center of mass of the experimental prototype to collect instantaneous angular velocity in the body coordinate system. A water tank experiment is conducted to generate a dataset. The collected pressure signals and gyroscope signals are preprocessed. Step 6: Fuse the preprocessed instantaneous angular velocity measurements from the gyroscope as supplementary observation information, and construct a residual objective function using the underwater vehicle's motion parameters and flow field environment parameters as parameters to be estimated, to perform attitude inversion on the underwater vehicle; Step 7: Compare the obtained underwater vehicle attitude inversion results with the actual attitude information, use relative root mean square error to evaluate the inversion accuracy, and complete the reliability verification of the above attitude inversion method.

2. The underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope as described in claim 1, characterized in that, In step 1, the underwater vehicle is a torpedo-type underwater vehicle. The spatial coordinate system includes three types of coordinate systems: the Earth fixed coordinate system, the navigation coordinate system, and the body coordinate system. The motion parameters of the underwater vehicle include attitude parameters in the navigation coordinate system, speed and angular velocity parameters in the body coordinate system, and flow field environment parameters are defined as flow field velocity parameters in the navigation coordinate system.

3. The underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope as described in claim 2, characterized in that, In step 2, the torpedo-shaped underwater vehicle adopts an axisymmetric structure. The pressure sensor extends along the axial direction of the torpedo-shaped underwater vehicle and is arranged circumferentially. The core layout parameters of the artificial side-line array include the axial spacing distance of the main array elements, the circumferential spacing angle of the head array elements, and the centroid offset distance.

4. The underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope as described in claim 3, characterized in that, In step 3, the quaternion model q By a rotation angle α and a unit rotation axis u The components of the unit rotation axis in the navigation coordinate system are ( u x , u y , u z In the navigation coordinate system, any attitude transformation of an underwater vehicle is equivalent to the body coordinate system rotating around a unit axis. u Rotation angle α Based on the above geometric relationships, step 3 includes: Step 3-1: Define the basic form of the quaternion model describing the vehicle's attitude: ,in: q 0 is the real part of the quaternion. q 1. q 2. q 3 is the imaginary part of the quaternion, sin is the sine function, and cos is the cosine function; Step 3-2: Derive the rotation matrix using the quaternion model C b n The matrix form of the quaternion model can be represented as: ； Step 3-3: Determine the coordinates M of any pressure sensor of the underwater vehicle in the body coordinate system. i,j b The coordinates M of the underwater vehicle in the navigation coordinate system after it rotates. i,j n The expression: .

5. The underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope as described in claim 4, characterized in that, Step 4 includes: Step 4-1: Based on the fact that the measured values ​​of any pressure sensor in the artificial lateral line array are the result of the superposition of hydrostatic pressure and multi-source dynamic pressure, the overall expression of the kinematic pressure signal model of the underwater vehicle is obtained as follows: ,in: p i,j For the first i The first axis, the first j Pressure measurement value from a circumferential pressure sensor p s,i,j For the first i The first axis, the first j The static pressure of a circumferential pressure sensor p d,i,j This is the sum of the dynamic pressure generated by the flow field, speed, and angular velocity at the pressure sensor. Step 4-2: Based on the fundamental properties of hydrostatics, and combining the Earth's fixed coordinate system and the navigation coordinate system, determine the water depth at which the underwater vehicle's mass point is located. d 0, and the coordinates of each sensor on the artificial lateral line in the navigation coordinate system, the hydrostatic model can be expressed as: ,in: p 0 represents atmospheric pressure. ρ The density of seawater, g For gravitational acceleration, (M i,j n )3 is the pressure sensor in the navigation coordinate system z The coordinate values ​​along the axis.

6. The underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope as described in claim 5, characterized in that, In step 5, a hydrodynamic experimental pool is selected to simulate the horizontal environmental flow field. Multi-dimensional experimental conditions are designed, and complex motion control of the experimental prototype is realized through a six-degree-of-freedom experimental platform. The experimental prototype is placed statically in the flow field, and static data is collected for sensor calibration. Simultaneously, artificial side-line array pressure signals and gyroscope angular velocity signals are collected to generate a dataset. The initial offset of the sensor is eliminated by using the mean of static pressure data as the baseline, and the inherent zero bias of the gyroscope is eliminated by using the mean of static angular velocity data as the zero bias value. The pressure signal and angular velocity signal are filtered separately to remove high-frequency turbulence noise and suppress random noise interference.

7. The underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope as described in claim 6, characterized in that, Step 6 includes: Step 6-1: Define the parameter vector to be estimated, with motion parameters and flow field parameters as the core. θ for: The preprocessed instantaneous angular velocity of the gyroscope is used as known observation information and directly fused into the multi-parameter kinematic pressure signal model constructed in step 4 as the input of the angular velocity dynamic pressure component. Step 6-2: Using the preprocessed artificial lateral line array pressure signal as the measured value and the multi-parameter kinematic pressure signal model that integrates gyroscope observation information as the theoretical value, construct the residual objective function. J ( θ ), represented as: Where: A and B are the number of axial and circumferential sensors in the artificial side-line array in step 2, respectively. p i,j ( t )for t Measured values ​​of artificial side-line array pressure signals after time-lapse preprocessing. p i,j ( θ The theoretical pressure value is obtained by substituting the parameters to be estimated into the multi-parameter kinematic pressure signal model. Step 6-3: Iterative process of attitude inversion based on LM algorithm, determining the initial values ​​of the parameters to be estimated. θ 0. Convergence threshold ε and initial value of damping coefficient μ 0. Calculate the theoretical pressure values ​​corresponding to all pressure sensors, and calculate the Jacobian matrix of the residual objective function with respect to each parameter to be estimated. J k Calculate the parameter correction amount according to the LM algorithm formula. Δθ k : ,in: k This represents the number of iterations in the LM algorithm. μ k The damping coefficient is... I It is the identity matrix. r k The residual vector; Step 6-4: Update the parameters to be estimated and calculate the updated residual objective function value. Determine whether the iteration has converged based on the convergence threshold. If it has not converged, adjust the damping coefficient based on the residual objective function value and continue iterating until the iteration converges. Extract quaternions from the optimal parameters to achieve attitude inversion.

8. The underwater vehicle attitude inversion method integrating artificial lateral line and gyroscope according to claim 7, characterized in that, In step 7: Using the actual attitude information from the pool experiment used to control the six-degree-of-freedom experimental platform as a benchmark, and converting it into quaternion form, the attitude inversion results output from step 6 are time-series matched with the actual attitude data. The relative root mean square error of the four attitude parameters of the quaternion model is used to quantify the attitude inversion accuracy of the underwater vehicle. The relative root mean square error per unit rotation axis of the quaternion model is expressed as... The relative root mean square error of the rotation angle of the quaternion model can be expressed as: , where: RRMSE( u ) represents the relative root mean square error (RRMSE) of attitude inversion per unit rotation axis. α ) represents the relative root mean square error of the rotation angle attitude inversion. t Sampling time, l The number of sampling points under experimental conditions. u 1( t )for t The rotation axis obtained by attitude inversion at any given time. u 0( t )for t Real-time rotation axis α 1( t )for t The rotation angle obtained from the attitude inversion at any given moment. α 0( t )for t Real-time rotation angle.

Images